Expression of the bipolar see-saw in Antarctic records during the last B. Stenni, D. Buiron, M. Frezzotti, S. Albani, C. Barbante, E. Bard, J.M. Barnola, M. Baroni, M. Baumgartner, M. Bonazza, et al.

To cite this version:

B. Stenni, D. Buiron, M. Frezzotti, S. Albani, C. Barbante, et al.. Expression of the bipolar see-saw in Antarctic climate records during the last deglaciation. Nature Geoscience, Nature Publishing Group, 2011, 4 (1), pp.46-49. ￿10.1038/ngeo1026￿. ￿insu-00647558￿

HAL Id: insu-00647558 https://hal-insu.archives-ouvertes.fr/insu-00647558 Submitted on 23 Feb 2021

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés.

Distributed under a Creative Commons Attribution| 4.0 International License Expression of the bipolar see-saw in Antarctic climate records during the last deglaciation

B. Stenni1, D. Buiron2, M. Frezzotti3*, S. Albani4, C. Barbante5, E. Bard6, J. M. Barnola2†, M. Baroni6, M. Baumgartner7, M. Bonazza1, E. Capron8, E. Castellano9, J. Chappellaz2, B. Delmonte4, S. Falourd8, L. Genoni1, P. Iacumin10, J. Jouzel8, S. Kipfstuhl11, A. Landais8, B. Lemieux-Dudon2, V. Maggi4, V. Masson-Delmotte8, C. Mazzola4, B. Minster8, M. Montagnat2, R. Mulvaney12, B. Narcisi3, H. Oerter11, F. Parrenin2, J. R. Petit2, C. Ritz2, C. Scarchilli3, A. Schilt7, S. Schüpbach7, J. Schwander7, E. Selmo10, M. Severi9, T. F. Stocker7 and R. Udisti9

Ice-core records of climate from Greenland and , separated by the (YD) cold event5,6. This show asynchronous temperature variations on millennial different sequence of events in the two hemispheres was related to timescales during the last glacial period1. The warming the thermal bipolar see-saw1,2. during the transition from glacial to interglacial conditions A prerequisite for studying the sequence and possible links was markedly different between the hemispheres, a pattern between climate events in Greenland and Antarctica is the attributed to the thermal bipolar see-saw2. However, a record determination of their relative timing with sufficient accuracy7. from the Ross Sea sector of East Antarctica has been suggested For sites located on the East Antarctic Plateau (EAP), high- 3 4 to be synchronous with Northern Hemisphere climate change . resolution CH4 records have been used to place the European Here we present a temperature record from the Talos Dome Project for Ice Coring in Antarctica (EPICA) Dronning Maud ice core, also located in the Ross Sea sector. We compare Land (EDML), and by extension the EPICA-Dome C (EDC) and our record with ice-core analyses from Greenland, based Vostok, Antarctic ice cores8, on the layer-counted Greenland Ice on methane synchronization4, and find clearly asynchronous Core Chronology 2005 (GICC05; see Supplementary Information). temperature changes during the deglaciation. We also find Although ice cores from the EAP show a coherent picture1,9,10, distinct differences in Antarctic records, pointing to differences coastal ice cores are expected to be influenced by regional in the climate evolution of the Indo-Pacific and Atlantic sectors signals related to the surrounding ocean. So far, few ice cores of Antarctica. In the Atlantic sector, we find that the rate from peripheral sites3,11 cover the last deglaciation. Critical of warming slowed between 16,000 and 14,500 years ago, for these coastal sequences is the relative depth of the last parallel with the deceleration of the rise in atmospheric deglaciation with respect to bedrock, where ice thinning and carbon dioxide concentrations and with a slight cooling deformation can perturb the stratigraphy. Furthermore, the over Greenland. In addition, our chronology supports the extremely high variability of strong wind scouring can induce hypothesis that the cooling of the Antarctic Cold Reversal is accumulation hiatuses. synchronous with the Bølling–Allerød warming in the northern The end of AIM1 as deduced from EAP ice cores is in phase hemisphere 14,700 years ago5. with the first rapid temperature change in Greenland5 (14.7 kyr bp). The period from about 8 to 25 kyr before present (bp) includes This synchronicity supports the thermal bipolar see-saw conceptual the climate transition from the last glacial to the Holocene. As model2, but is in apparent contradiction with the timing of climate documented from polar ice cores and other climate archives, the change shown by the Taylor Dome (TYD) ice core in the Ross Sea pattern of climate changes throughout this transition is different sector3. The timing of the stable isotope record from this coastal ice between Antarctica and the surrounding Southern Ocean and the core seemed more similar to Greenland rather to Antarctic cores, Northern Hemisphere. The steady Antarctic deglacial warming with a first warming ending abruptly at ∼14 kyr bp (ref. 3), although reaches a first maximum (Antarctic Isotopic Maximum AIM1; this dating was already questioned12. Another coastal ice core, Law ref. 1) followed by an interruption towards cooler conditions during Dome (LD, East Antarctica), conversely, ends the first warming at the Antarctic Cold Reversal (ACR). Conversely, Greenland records ∼15 kyr bp, suggesting that the ACR does not follow the abrupt show two rapid-warming phases at the onset of the Dansgaard– warming of DO1 (ref. 11). However, in this critical time interval, Oeschger-1 (DO1) event (Bølling–Allerød interstadial, B/A) and the questions regarding the integrity of the timescale (TYD) and abrupt

1Department of Geosciences, University of Trieste, 34127 Trieste, Italy, 2Laboratoire de Glaciologie et de Géophysique de l’Environnement (CNRS-Université Joseph Fourier - Grenoble), 38402 Saint Martin d’Hères cedex, France, 3ENEA, CR Casaccia, 00123, Roma, Italy, 4Environmental Sciences Department, University of Milano Bicocca, 20126 Milano, Italy, 5Department of Environmental Sciences, University Cà Foscari of Venice, and IDPA-CNR, 30123 Venezia, Italy, 6CEREGE, UMR 6635 CNRS, IRD, University Aix-Marseille, Collège de France, Europôle de l’Arbois, BP80, 13545 Aix-en-Provence cédex 4, France, 7Climate and Environmental Physics, Physics Institute, University of Bern, 3012 Bern, Switzerland, 8Laboratoire des Sciences du Climat et de l’Environnement (IPSL/CEA-CNRS-UVSQ UMR 8212), CEA Saclay, 91191 Gif-sur-Yvette cédex, France, 9Department of Chemistry, University of Firenze, 50019 Sesto Fiorentino, Italy, 10Department of Earth Sciences, University of Parma, 43100 Parma, Italy, 11Alfred-Wegener-Institute for Polar and Marine Research, 27568 Bremerhaven, Germany, 12British Antarctic Survey, NERC, Cambridge, CB3 0ET, UK. †Deceased. *e-mail: [email protected].

1 a CH ¬34 Holocene Interval studied MIS5.5 ) 700

MIS7.5 4 TALDICE 600 (ppbv) OW ¬36 AIM1 Greenland composite 500 SM ¬38 b ¬34 DO0 DO1 400 ( ¬36 YD H1 DO2 H2 O ¬40 ACR LGMM 18 H0

¬38 δ SMOW NGRIP DICE ¬42 ¬40

0 200 400 600 800 1,000 1,200 1,400 1,600 O ( ¬42 ) 18 Depth (m) δ c ¬44 Figure 1 | The TALDICE stable isotope profile (δ18O) versus depth. The 6 GGC5 231

interval studied (8–25 kyr BP) is marked with grey shading. MIS 5.5, 7.5, Pa/ 7

LGM, AIM1 and ACR are indicated. The satellite image of Antarctica shows 230

the location of the TALDICE ice core. 8 Th d 260 9 × 10¬2 reduction in accumulation and thus limited time resolution (LD) 240 (ppmv) render some earlier conclusions questionable. 2 220

Here we investigate a new 1,620-m-deep ice core drilled at CO 200 EDC CO2 Talos Dome (TLD) in the framework of the Talos Dome Ice Core ¬44 e (TALDICE) project (www.taldice.org). TLD is a peripheral dome of ) ( ¬46

East Antarctica, located in the Ross Sea sector (Fig. 1). The moisture SMOW EDML AIM2 sources for TLD are mainly located in the Pacific and Indian sectors f ¬50 ¬48 of the Southern Ocean (see Supplementary Information). The MWP1a whole TALDICE ice core provides a palaeoclimate record covering ¬52 ¬50 the past 250 kyr back to Marine Isotope Stage (MIS) 7.5 (Fig. 1). SMOW ¬54 ¬52 The TALDICE oxygen-isotope (δ18O, a proxy of local temperature)

O ( O ( ¬56 ) record is presented here for the entire core (Fig. 1) and for the 18 EDC bp δ time window between 8 and 25 kyr (Fig. 2). TALDICE enables ¬58 AIM0 decadal-scale resolution during the last deglaciation owing to its g AIM1 ¬36 relatively high accumulation rate (80 kg m−2 yr−1). In this work, ) ( ACR the TALDICE-1 chronology (see Supplementary Information) ¬38 SMOW 8 has been set up using a new inverse method , generating an TALDICE ¬40 optimal compromise between an a priori scenario set up from a glaciological model and chronological information from different ¬42 4 time markers. CH4 data are used to synchronize TALDICE to 8 10 12 14 16 18 20 22 24 the North Greenland Ice Core Project (NGRIP) ice core on the Age scale kyr BP (AD 1950) GICC05 age scale (see Supplementary Information), providing a relative synchronization error less than 100 yr for the sharp CH4 Figure 2 | Compilation of palaeoclimatic records from ice and marine transitions. TALDICE is the first coastal site with glaciological cores to depict the bipolar sequence of events during the last termination. 4 18 characteristics enabling accurate dating during the entire last a, CH4 records of Greenland composite and TALDICE (this study); b, δ O deglaciation, which starts at around 800 m depth (∼50% of the record from NGRIP (ref. 6); c, 231Pa/230Th record of marine core GGC5, total ice thickness; see Supplementary Information). The resulting from Bermuda rise in the deep western subtropical Atlantic, taken as a 19 chronological uncertainty is ±300 yr between 8 and 15 kyr, ±500 yr proxy for Atlantic meridional overturning circulation strength ; d, CO2 back to 17.5 kyr and up to ±1.5 kyr during the glacial period. from EDC (ref. 18); e, δ18O record from EDML (ref. 1); f, δ18O record from The pattern of δ18O during deglaciation is similar at TALDICE EDC (ref. 10); g, δ18O record from TALDICE (this study). EDML, EDC and and EDC (Fig. 2), despite their different geographical positions and TALDICE are synchronized on the GICC05 timescale using a new inverse moisture sources (see Supplementary Information). The transition method. GGC5 data are shown on their own radiocarbon timescale. The starts synchronously at 18.2 ± 0.7 kyr bp in TALDICE, EDML, dotted lines correspond to the ramps obtained with the Rampfit and EDC and Dome Fuji (DF) records (Figs 2 and 3), a few kiloyears Breakfit software. The YD and DO interstadials are indicated5,6. The black before the Antarctic ice-sheet margin retreat13. Warming appears horizontal bar corresponds to MWP1a (ref. 22); the arrows represent the coeval with the ages reported for mid-latitude glacier retreat ages of Heinrich layers 1 and 2. The triangles indicate synchronization CH4 from both hemispheres14 and the rapid sea-level rise at 19 kyr bp tie points. The grey vertical bars correspond to the AIM2 event, the start of (ref. 13). The first part of deglaciation culminates in the AIM1 the deglaciation (18.2±0.7 kyr BP), the slowing of the warming at EDML event at 14.7±0.3 kyr bp, followed by the ACR cooling (Fig. 2) until (16.0±0.2 kyr BP), the AIM1 event (14.7±0.3 kyr BP), the end of the ACR 12.7 ± 0.3 kyr bp, and by the final warming towards the onset of (12.7±0.3) and the AIM0 (11.9±0.3) event as inferred from δ18O ice-core Holocene. The ACR cooling appears with weaker δ18O amplitude records (this study). in TALDICE than in EDC. TALDICE confirms that the TYD chronology3 is almost at EDC. The lack of an AIM2 signal in the Indo-Pacific sector certainly incorrect12 and enables us to refute that the deglacial is confirmed by the TALDICE data (Fig. 2). A specific deglacial history in the Ross Sea area was synchronous with the Northern pattern is observed in EDML–DF ice cores compared with EDC– Hemisphere (Figs 2 and 3). TALDICE. These two last cores show a ∼3.5 kyr warming at an During the (LGM) and the deglaciation, almost constant rate (1 and 1.3h δ18O kyr−1, respectively) peaking the comparison of δ18O records around Antarctica highlights in AIM1, whereas EDML–DF depict a reduced warming rate (about regional features. Earlier studies revealed a smoother shape of 0.5h δ18O kyr−1) between 16.0 ± 0.2 kyr bp and AIM1 (Figs 2 glacial AIM events at EDML compared with EDC (ref. 15) as and 3). This slowdown of the warming can be also depicted from well as an identification of AIM2 at EDML–DF (refs 1,9) but not EDC–TALDICE records but is negligible with respect to EDML–DF.

2 ) ) ) C

° 16 ( ¬44 OW 14 OW ¬52 ¬46 SM SM ¬54 12 ¬48 ¬56 ( 10 ¬50 ( Dome Fuji (DF) O O ¬58 8 lkenone SST 1 ODP 1233 41° S 74.5° W EDML 75° S 0° 18 77.5° S 37.5° E

¬52

A δ δ 8 10 12 14 16 18 20 22 24 8 10 12 14 16 18 20 22 24 8 10 12 14 16 18 20 22 24 Age kyr BP (AD 1950) Age kyr BP (AD 1950) Age kyr BP (AD 1950) ) 445° S LGM ice-sheet reconstruction W N O ¬22 OCEA M LGM winter sea-ice extension IC S ¬24 NT A 60° S L ( ¬26 LGM summer sea-ice extension T A Law Dome (LD) I O N

18 ¬28 ° ° D 66.8 S 112.8 E Modern winter sea-ice extension dd ll δ I EDM A N 8 10 12 14 16 18 20 22 24 Modern summer sea-ice extension ODODPD 1231233 DF O Age kyr BP (AD 1950)

C

S E ° ° ° 90 W 90 E )

) BY A 75 N

VS W ¬48 W

TYDY O

O ¬34

LDLD ¬50

EDCDC M M N

¬36 S S

A Ross TLD ¬52

E

¬38 C Sea (

O

( ¬54

C

¬40 I O

O F

° ° I 8 ° ° 1

Byrd (BY) 80 S 120 W C ¬56 EDC 75 S 123.3 E 18 A P δ ¬42 δ 8 10 12 14 16 18 20 22 24 8 10 12 14 16 18 20 22 24 Age kyr BP (AD 1950) MD97-212022120 Age kyr BP (AD 1950) 1880° ) ) W C) ¬36 ° W O (

14 O

¬36 M ¬38 S 12 M S ¬38 ¬40 10 ( ¬40 ¬42

( TALDICE (TLD) Taylor Dome (TYD)

8 O 8 O MD97-2120 45.5° S 175° E ¬42 72.8° S 159.2° E 1 ¬44 77.8° S 158.7° E 8

lkenone SST 6 1 δ

A 8 10 12 14 16 18 20 22 24 δ 8 10 12 14 16 18 20 22 24 8 10 12 14 16 18 20 22 24 Age kyr BP (AD 1950) Age kyr BP (AD 1950) Age kyr BP (AD 1950)

Figure 3 | Compilation of Southern Hemisphere palaeoclimatic records from ice and marine-sediment cores during the last deglaciation. From top left in clockwise order, alkenone-SST record from ODP 1233 (ref. 16); δ18O record from EDML (ref. 1), DF (ref. 9), LD (ref. 11), EDC (ref. 10), TYD (ref. 3) and TALDICE (this study); alkenone-SST record from MD97-2120 (ref. 17); δ18O record from Byrd28. EDML, EDC and TALDICE are synchronized on the GICC05 timescale using a new inverse method. DF, LD, TYD and Byrd data are shown on their own timescale. ODP 1233 data are shown on its timescale16 and MD97-2120 on the timescale developed by ref. 29. The map also shows, with circles, the locations of the ice cores of the Antarctic Ice Sheet and sediment cores in the Southern Ocean cited in the text (VS: Vostok). In blue, a sketch of the LGM ice-sheet reconstruction. The blue and light-blue lines show LGM and modern winter (dashed lines) and summer (dotted lines) sea-ice extension30, respectively.

Albeit hampered by the lack of coherent age scales, the comparison moisture-source temperatures are derived from deuterium-excess of the deglaciation δ18O records obtained on various ice cores data15 at EDML and EDC. Different shifts in moisture sources (Fig. 3) reveals regional site-specific patterns in terms of warming are probably linked to reorganization of atmospheric circulation rates and magnitudes. A change in the rate of warming around at basin scale. A southward displacement of westerly winds has 16.7 kyr bp is also observed in an alkenone-sea surface temperature recently been reported21 during Northern Hemisphere stadials. (SST) record at the ODP 1233 site near the Chilean coast16, The different rates of warming between 16.0 kyr bp and AIM1, although no further slope changes corresponding to AIM1 or ACR observed at EDML–DF and EDC–TALDICE, suggest a different are shown. Conversely, the alkenone-SST record from a marine expression of the bipolar see-saw in the two regions. The underlying core (MD97-2120) collected in the Southwest Pacific17 shows an mechanisms will be investigated in future through new observations uninterrupted deglacial warming comparable to EDC–TALDICE and modelling studies. records (Fig. 3). Therefore, both marine and ice core sequences The onset of ACR in TALDICE and EDC chronologies depict a different rate of warming in the South Chilean coast/South corresponds to the DO1 onset (14.7 kyr bp) in the NGRIP ice core5 Atlantic compared with the Indo/South Pacific, starting around and with the minimum 1 age chronology at LD (ref. 11). The ACR 16 kyr bp. The warming slowdown at EDML–DF is synchronous onset also corresponds to the timing of the exceptionally large Melt 18 with the deceleration of the CO2 rise , with the unusually high Water Pulse 1a (MWP1a) inferred from relative sea-level records. 231Pa/230Th ratio in abyssal North Atlantic sediments interpreted MWP1a occurred at around 14.6 kyr bp with a global sea-level as a weaker Atlantic meridional overturning circulation strength19, rise between 10 and 20 m in a few centuries22. Observational with minimum δ18O values at NGRIP (Fig. 2) and with distinct data and modelling studies indicate that Antarctica contributed change in the strength of the Asian monsoon20. Specific changes partly to MWP1a (ref. 13). Our new synchronized isotopic records in the Atlantic sector linked with atmospheric circulation (westerly suggest that the ACR could be a response to MWP1a, which in wind position), winter sea-ice extension and Antarctic Bottom turn may partially originate from Antarctica after a prolonged Water formation may be involved in the parallel deceleration of deglacial warming. EDML and TALDICE are located close to the CO2 rise. Although it is not clearly imprinted in the Atlantic the most important areas of Antarctic deepwater formation. A sector (EDML–DF), the culmination of AIM1 is clearly visible massive freshwater release in the high-latitude Southern Ocean in the Indian Ocean sector dated at 14.7 ± 0.3 kyr bp on the is expected to shut down convection, reduce southward heat TALDICE-1 chronology, 14.5 ± 0.2 kyr bp on the new EDC age transport and increase sea ice and associated albedo feedbacks, scale8 (Fig. 2). Between 16.0 kyr bp and AIM1, different trends of inducing high-latitude Southern Ocean cooling23. This effect is still

3 a subject of debate between modelling groups7,24, which reinforces 14. Schaefer, J. M. et al. Near-synchronous interhemispheric termination of the the importance of palaeoclimate records as targets of numerical Last Glacial Maximum in mid-latitudes. Science 312, 1510–1513 (2006). simulations. Our data support the hypothesis that the DO1 (B/A) 15. Stenni, B. et al. The deuterium excess records of EPICA Dome C and Dronning Maud Land ice cores (East Antarctica). Quat. Sci. Rev. 29, 146–159 (2010). is due to a sudden Atlantic meridional overturning circulation 16. Lamy, F. et al. Modulation of the bipolar seesaw in the Southeast Pacific during 7 recovery from the Heinrich 1 event and by a decrease in Antarctic Termination 1. Earth Planet. Sci. Lett. 259, 400–413 (2007). Bottom Water formation in competitive see-saw behaviour with 17. Pahnke, K. & Sachs, J. P. Sea surface temperatures of southern midlatitudes North Atlantic Deep Water25. 0–160 kyr B.P. Paleoceanography 21, PA2003 (2006). There is now growing evidence that the end of the YD in 18. Monnin, E. et al. Atmospheric CO2 concentrations over the last glacial termination. Science 291, 112–114 (2001). the North shows the same characteristics of a typical abrupt DO 19. McManus, J. F., Francois, R., Gherardi, J-M., Keigwin, L. & Brown-Leger, S. warming event (DO0), associated with the early Antarctic Holocene Collapse and rapid resumption of Atlantic meridional circulation linked to optimum (AIM0) in a thermal bipolar see-saw pattern26. The deglacial climate changes. Nature 428, 834–837 (2004). ACR–AIM0 transition starts synchronously at 12.7–12.9 kyr bp in 20. Broecker, W. S. et al. Putting the Youger Dryas cold event into context. Quat. Sci. Rev. 29, 1078–1081 (2010). TALDICE–EDC–EDML (see Supplementary Information), syn- 328, 27 21. Denton, G. H. et al. The last glacial termination. Science 1652–1656 (2010). chronously with the precisely dated glacier retreat in New Zealand , 22. Deschamps, P. et al. Synchroneity of Meltwater Pulse 1A and the Bølling onset: demonstrating a coherent picture for the end of deglaciation New evidence from the IODP ‘Tahiti Sea-Level’ expedition. Geophys. Res. Abstr. between mid- and high latitudes of the Southern Hemisphere. 11, EGU2009-10233 (2009). The weaker δ18O amplitude of ACR and AIM0 at TALDICE, 23. Weaver, A. J., Saenko, O. A., Clark, P. U. & Mitrovica, J. X. Meltwater Pulse 1A from Antarctica as a trigger of the Bølling–Allerød warm interval. Science 299, compared with other EAP ice cores (Fig. 2), can be explained 1709–1713 (2003). with local elevation and air-mass-trajectory changes (see Supple- 24. Knorr, G. & Lohmann, G. Southern Ocean origin for the resumption of Atlantic mentary Information). thermohaline circulation during deglaciation. Nature 424, 532–536 (2003). Located in the Ross Sea sector, TLD is the most remote ice- 25. Pahnke, K., Goldstein, S. L. & Hemming, S. R. Abrupt changes in Antarctic core site from the North Atlantic ‘centre of action’, and still intermediate water circulation over the past 25,000 years. Nature Geosci. 1, 870–874 (2008). clearly shows a thermal bipolar see-saw behaviour during the last 26. Masson-Delmotte, V. et al. Abrupt change of Antarctic moisture origin at the deglaciation. This paper highlights the distinct regional climate end of Termination II. Proc. Natl Acad. Sci. USA 107, 12091–12094 (2010). fingerprint between the Indo-Pacific and the Atlantic sectors 27. Kaplan, M. R. et al. Glacier retreat in New Zealand during the Youger Dryas between 16.0 kyr bp and AIM1. Understanding and simulating the stadial. Nature 467, 194–197 (2010). bipolar and peri-Antarctic sequences and patterns of events and 28. Blunier, T. & Brook, E. J. Timing of millennial-scale climate change in Antarctica and Greenland during the last glacial period. Science 291, their interplay with the carbon cycle are critical challenges for earth 109–112 (2001). and climate system modelling, although many gaps remain in the 29. Rose, K. A. et al. Upper-ocean-to-atmosphere radiocarbon offsets imply fast description of the past regional Antarctic climate variability. deglacial carbon dioxide release. Nature 466, 1093–1097 (2010). 30. Gersonde, R., Crosta, X., Abelmann, A. & Armand, L. Sea-surface temperature and sea ice distribution of the Southern Ocean at the EPILOG Last Glacial Maximum—a circum-Antarctic view based on siliceous microfossil records. Quat. Sci. Rev. 24, 869–896 (2005). References 1. EPICA Community Members. One-to-one coupling of glacial climate Acknowledgements variability in Greenland and Antarctica. Nature 444, 195–198 (2006). We thank the logistic and drilling TALDICE team. We thank M. Mudelsee for discussion 2. Stocker, T. F. & Johnsen, S. J. A minimum thermodynamic model for the regarding the use of the Rampfit software (www.mudelsee.com). The Talos Dome Ice bipolar seesaw. Paleoceanography 18, 1087 (2003). Core Project (TALDICE), a joint European programme led by Italy, is funded by 3. Steig, E. J. et al. Synchronous climate changes in Antarctica and the North national contributions from Italy, France, Germany, Switzerland and the United Atlantic. Science 282, 92–95 (1998). Kingdom. The main logistical support was provided by Programma Nazionale di 4. Blunier, T. et al. Synchronization of ice core records via atmospheric gases. Ricerche in Antartide (PNRA) at Talos Dome. This is TALDICE publication no 9. A file Clim. Past 3, 325–330 (2007). containing the new ice core δ18O record from TALDICE can be downloaded from the 5. Steffensen, J. P. et al. High-resolution Greenland ice core data show abrupt website of the World Data Center for Paleoclimatology (www.ncdc.noaa.gov/paleo) and climate change happens in few years. Science 321, 680–684 (2008). from the TALDICE website (www.taldice.org). 6. North Greenland Ice Core Project Members. High resolution climate record of the northern hemisphere reaching into the last interglacial period. Nature 431, Author contributions 147–151 (2004). B.S., D.B., M. Baumgartner, M. Bonazza, S.F., L.G., V.M-D., B.M., A.S., S.S. and E.S. 18 7. Liu, Z. et al. Transient simulation of last deglaciation with a new mechanism contributed to new TALDICE δ O and CH4 measurements. D.B., S.A., E.B., M. Baroni, for Bølling–Allerød warming. Science 325, 310–314 (2009). E. Capron, E. Castellano, B.D., A.L., B.L-D., C.M., M.M., B.N., F.P., J.R.P., C.R., M.S. 8. Lemieux-Dudon, B. et al. Consistent dating for Antarctic and Greenland ice and R.U. contributed to measurements and modelling used for TALDICE dating and ice cores. Quat. Sci. Rev. 29, 8–20 (2010). flow description. B.S. and M.F. prepared the figures and wrote most of the manuscript 9. Watanabe, O. et al. Homogeneous climate variability across East Antarctica with the main contributions of D.B., J.C., B.D., V.M-D., J.J., B.L-D., B.N., J.R.P. and over the past three glacial cycles. Nature 422, 509–512 (2003). T.F.S. All authors worked on the comparison of the different records and their 10. Jouzel, J. et al. Orbital and millennial Antarctic climate variability over the past interpretations. 800,000 years. Science 317, 793–796 (2007). 11. Morgan, V. et al. Relative timing of deglacial climate events in Antarctica and Additional information Greenland. Science 297, 1862–1864 (2002). The authors declare no competing financial interests. Supplementary information 12. Mulvaney, R. et al. The transition from the last glacial period in inland and accompanies this paper on www.nature.com/naturegeoscience. Reprints and permissions near-coastal Antarctica. Geophys. Res. Lett. 27, 2673–2676 (2000). information is available online at http://npg.nature.com/reprintsandpermissions. 13. Clark, P. U. et al. The Last Glacial Maximum. Science 325, 710–714 (2009). Correspondence and requests for materials should be addressed to M.F.

4